analyzing transit travel time performanceonlinepubs.trb.org/onlinepubs/trr/1983/915/915-001.pdf ·...
TRANSCRIPT
I
I
Transportation Research Record 915 1
Analyzing Transit Travel Time Performance
HERBERT S. LEVINSON
A detailed analysis of transit speeds, delays, and dwell times based on surveys conducted in a cross section of U.S. cities is summarized. The relationships and parameters provide inputs for planning service changes and assessing their impacts. The surveys and analyses find that car speeds are consistently 1.4 to 1.6 times as fast as bus speeds; time the typical bus speeds about48 to 75 percent of its moving, 9 to 26 percent at passenger stops, and 12 to 26 percent in traffic delays; and peak-hour bus travel times approximate 4.2 min/mile in suburbs, 6.0 in the city, and 11.50 in the central business district. Bus dwell times (including door opening and closing) approximate 5 sec plus 2. 75 times the number of passengers; during peak hours local buses stop at 68 to 78 percent ofthe designated stops. Bus travel times and speeds were derived as a function of stop frequency, stop duration, and bus acceleration and deceleration times observed in the field. Reducing bus stops from eight to six per mile and dwell times from 20 to 15 sec would reduce travel times from 6 to 4.3 min/mile, a time saving greater than that which could be achieved by eliminating traffic congestion. Transit performance should be improved by keeping the number of stopping places to a minimum. Fare-collection policies and door configurations and widths are important in reducing dwell time, especially along highdensity routes. Such time savings will likely exceed those achieved from providing bus priority measures or improving traffic flow.
Transit travel times and operating speeds influence service attractiveness, costs, and efficiency. They also provide important descriptions of system performance for use in the transportation planning process. Yet, despite their importance, relatively few studies have been made to quantify these factors as they relate to ridership density, traffic conditions, and land use.
A detailed analysis of transit speeds, delays, and dwell times based on surveys conducted in a cross section of u.s. cities in 1980 (1) is summarized. The study was initially designed to-verify and update !NET reports on transit speed and roadway type (~).
In a broader sense, however, it provides parameters for use in planning service changes and estimating their impacts.
The study included the following steps:
1. Available literature on transit-delay characteristics over the last several decades was assembled and analyzed;
2. Field studies were conducted of bus (and rail) performance in Boston, Chicago, New Haven, and San Francisco in 1979 and 19801
3. Transit acceleration and deceleration characteristics were simulated and compared with actual times observed in the field; and
4. Results were integrated to produce a consistent and rea~istic picture of transit performance in U.S. cities.
Table 1. Comparative bus and car speeds for selected urban areas.
Relation of Car to Bus Speed (min/mile)
Morning Peak Midday No. of
City and Year Routes Ratio SD Ratio SD
Chicago (Loop), 1950 NA 1.39° Dallas, 1972 14 1.61 0.28 New Haven, 197 5 15 1.54 0.22 Midtown Manhattan, 1968 16 1.59 0.35 1.63 0.43 San Jose, 1968 NA 1.42 1.48
Notes: NA= not available.
Research findings address the following areas:
1. A comparison of line-haul bus and car travel times,
2. Bus speeds and delay, 3. Passenger service times at bus stops, 4. Bus (and train) dwell times (per stop) and
stop frequencies (stops per mile), 5. Bus acceleration and deceleration, and 6. Transit speeds as a function of stop frequen
cies and dwell times.
The components of transit travel time that ham~ been quantified are shown in Figure 1.
BUS AND CAR SPEEDS
Ratios of car to bus speed in Chicago's Loop, midtown Manhattan, Dallas, New Haven, and San Jose are shown in Table 1 (3-6). Car speeds are consistently 1.4 to 1. 6 times -faster than bus speeds. These ratios appear independent of year of study or type of city.
TRANSIT SPEED AND DELAY
Peak-hour transit speed and delay data for eight cities are summarized in Tables 2 and 3 <l-!Q.l • Minutes per mile (delay rate) has been used as the basic parameter, since it enables times to be added as needed. Means, standard deviations, and percentage distributions are given for time spent moving, at passenger stops, and in traffic delays. (The
Figure 1. Transit time components.
UJ 0 z < tUJ l5
A
TIME BETWEEN STOPS
Evening Peak
Ratio SD
1.38 1.65 0.16
1.48 0.30 1.37
A-STOP B-ACCELERATE C-CRUISE D-DECELERATE
Some data are from the Bureau of Traffic Operations, New York City Department of Transportation. 8 8:00 a.m. to 6:00 p.m.
Transportation Research Record 915
Table 2. Travel time and delay for typical transit routes.
Proportion of Journey Time(%) Spent for
No. of Avg. Travel Routes or Time Traffic Passenger
Mode City and Year Streets (min/mile) Delays Stops Moving Remarks
Bus Oakland, Alameda-Contra 4 4.95 19.4 26.7 53.9' Suburban Costa County, CA; 1979 1 3.18 18.6 23.6 57.9 Intercity
Minneapolis, MN; 1977 3' 11.34 25.8 24.0 50.2 CBD Philadelphia, PA; 1977 2• 11.41 26.5 25.8 47.7 CBD Santa Clara, CA; 1969 3 4.38 16.2 9.1 74.7 Suburban St. Louis, MO; 1957-1958 20 5.47 12.1 17.9 70.0 City lines New Haven, CT; 1979-1980 2 6.14 19.0 18.4 62.6 Urban-suburban
Streetcar Beacon St., Boston, MA; 1968 1 6.06 14.8 22.9 62.3 In center reservation St. Louis, MO; 1957-1958 4 6.60 12.7 17.7 69.6 City lines
Note: Some data are from field surveys in conjunction with the Regional Planning Agency of South Central Connecticut. 8 Streets.
Table 3. Transit travel times for typical routes.
Travel Time (min/mile)
No. of Avg or Total Traffic Delays Passenger Stops Moving Routes or
Mode City and Year Streets Mean SD Mean SD Mean SD Mean SD Remarks
Bus Oakland, Alameda-Contra 4 4.95 0.37 0.96 0.19 1.32 0.26 2.67 0.16 Suburban Costa County, CA; 1979 1 3.18 0.25 0.59 0.11 0.75 0.07 1.84 0.26 Intercity
Minneapolis, MN; 1977 3 11.34 1.96 2.93 1.46 2.72 1.23 5.69 1.19 CBD New Haven, CT; 1979-1980 1 5.88 0.51 0.99 0.14 1.15 0.22 3.74 0.31 Urban
1 6.40 0.86 1.35 0.38 1.10 0.37 3.95 0.73 Urban Philadelphia, PA; 1977 2 11.41 0.88 3.03 0.50 2.94 0.64 5.44 0.36 CBD Santa Clara, CA; 1969 3 4.38 0.20 0.71 0.08 0.40 0.06 3.27 0.10 Suburban (low
density) St. Louis, MO; 1957-1958 20 5.47 0.48 0.66 0.29 0.98 0.21 3.83 0.37 City lines
Streetcar Beacon St., Boston, MA; 1968 6.07 0.83 0.90 0.24 1.39 0.46 3.78 0.22 In center reservation
St. Louis, MO; 1957-1958 4 6.60 1.09 0.84 0.46 1.17 0.24 4.59 0.38 City lines
Note: Some data are from field surveys in conjunction with the Regional Planning Agency of South Central Connecticut.
Table 4. Estimated peak-hour transit travel time~ by component.
Travel Time (min/mile)'
Component CBD City Suburbs
Traffic delay 3.00 ± 1.00 0.90 ± 0.30 0.70±0.10 Passenger stops 3.00 ± 1.00 1.20 ± 0.30 0.50 ± 0.10 Moving 5.50 ± 1.00 3.90 ± 0 30 3.00±0.12 Total 11.50 ± 3.00 6.00 ± 0.90 4.20 ± 0.30
Note: Data are from Tables 2 and 3. 8 Plus-or-minus values represent one standard deviation,
standard deviations reflect the variations among average times reported for various bus or streetcar routes in each community.)
Reported ranges for U.S. cities in the time spent enroute are moving, 46 to 75 percent; u.t passenger stops, 9 to 26 percent; and in traffic delays, 12 to 26 percent.
Transit travel times vary by type and location of route. Generalized peak-hour travel times for the central business district (CBD), central city, and suburban bus lines by time component are shown in Table 4 and Figure 2 in minutes per mile. The following characteristics may be noted:
1. Peak-hour bus travel times approximate 4. 20 min/mile in suburban areas, 6. 00 min/mile in the central city, and 11.50 min/mile in the CBD.
"· The time in motion approximate s 3.00 min/mile in the suburbs, 3. 90 min/mile in the central city,
Figure 2. Peak-hour bus travel times.
TRAFFIC DELAY
PASS. STOPS
MOVING
11.5
6.0
4.2
C.B.D. CITY SUBURBS
and 5. 50 min/mile in the CBD. inversely with the frequency of
3. Passenger stops account the suburbs, 1.20 min/mile in min/mile in the CBD.
It appears to vary stops. for O. 50 min/mile in
the city, and 3.00
4. Traffic delay amounts to 0.70 min/mile in the suburbs, 0.90 min/mile in the city, and 3.00 min/mile in the CBD.
In the central city, passenger stop delay exceeds
Transportation Research Record 915
traffic delays, whereas they are about equal in the CBD. Therefore, ways to reduce passenger delays on a citywide basis may prove more beneficial than efforts focused only on alleviating traffic congestion at key locations.
BUS-STOP FREQUENCIES AND PASSENGER SERVICE TIMES
Information on passenger stops and dwell times was obtained from specially conducted field surveys in Boston, Chicago, New Haven, and San Francisco. The results of these studies are summarized in Table 5 in which the following information is given on a route-by-route basis:
1. Route location and distance; 2. Range and mean of actual stops made per mile; 3. Range and mean for maximum dwell times re
ported; 4. Range and mean for the average dwell times
reported; and 5. Representative formulas for estimating pas
senger dwell times, including time spent opening and closing doors.
Table 5. Summary of observed bus-stop frequencies and passenger dwell times.
City
Boston
Chicago
New Haven
San Francisco
Boston
Bus Route
71
71
77
77
240A
11 11
22
B-1 B-1
Dl-2
Dl-2
Jl,2,3
Jl,2,3
Distance (miles)
4.0
4.0
3.3
3.3
5.3
5.3
8.8
6.2 1.2
1.7
4.1 8.2
6.6
10.4
4.8
8.5
Q 3.0
Stockton NA
Green Line
Green Line
2.3
2.3
Time of Day
p.m.
Midday
a.m.
Midday
p.m.
Midday
Midday
a.m. a.m.
Midday
p.m. p.m.
p.m.
p.m.
iJ.m.
p.m.
a.m.
a.m.
a.m.
p.m.
Stops per Mile
Direction
Both
Both
SB
Both
NB
Both
Both
SB SB
SB
SB
Range"
5.0-5.5
3.3-3.5
6.7
2.7-4.5
4.7
1.3-1.5
1.1-1.5
5.2 6.7
6.5
Southern leg 3.2-4.7 SB through 5.1-7.2 center
NB (northern 4.1-5.6 leg)
SB through 3.7-5.0 center
SB (southern 5.2-6.3 leg)
NB through 2.9-5.5 center
EB To center
Inbound
Outbound
5.3-6.0
6.2
6.6
3
7
3
s 7
6
4 6
4
6
4
6
6
6
3
The formulas take the following form:
T = an + b
where n is the number of interchanging passengers per bus and T is the total stopped time per bus in seconds. Representative values of the coefficients a and b are as follows:
Location Activity ~ ~ Boston Mainly discharging 1.2-1.7 4.0 Boston Paying when outbound 2.0 New Haven Boarding and alighting 2.6-3.0
4.3 3.9-5. 6
The formula T = 2.75n + 5 sec provides a reasonable estimate of the dwell times in any community.
The variations in dwell time along specific routes reflect the location of stop, surrounding land uses, and the number of interchanging bus lines. Although stops generally average less than 20 sec, buses spent 30 to 60 sec at major transfer points, terminals, or rail-bus interchange locations. Examples of dwell times at major bus stops are shown in Table 6.
In estimating bus performance, it is necessary to
Dwell Time per Stop (sec)
Maximum
Range" Mean
29-35
37-61
37
12-29
38
21-28
34-54
40 40
27
21-40d 40-51 35.53d
26-39d
25-53d
30-45d
22-32d
9-2od
30
NA
NA
32
49
37
20
38
25
41
40 40
27
33
37
Average
Range'
10.4-13.2
14.4-17.3
l l.6c
8.6-13.0
13.1
7 .0-8.5c
9.8-18.2c
13.2 17.7
14.1
8.5-14.3d
l J.0-15.7d
10.l-13.5d
11.3-13.ld
9.5-11.Sd
8.7-l l.9d
4.5-8.2d
21
NA
NA
Mean
11.8
Representative Formula
l.7n + 4.0
15.7 2.5n+5.0
1 l.6c 2.6n + 2.1
10.9 3.ln + 5.1
13. l 2.0n + 4.3
7.8° l.2n+4.0
13.6° 3.7n + 5.7c
13.2 NA 17.7
14.1 NA
3.2n + 3.9 14.5d 2.7n + 5.6
l l.6d 2.5n + 5.1
ll.9d 3.0n+5.l
10.8d 2.8n + 4.4
10.7d 2.6n + 4.6
6.5d NA
21.0 NA
18.0 NA
17.5 NA
Remarks
Urban route, mainly discharging passengers
Urban route, high density
White-collar passengers
Urban-suburban route
Suburban-urban route
Pay when entering inbound, when leaving outbound; suburban limited stops
Mainly alighting passengers; suburban limited stops
Suburban line
Urban line Urban line, heavy section
Central section, high-density line
Urban route Urban route
Urban-suburban route
Urban-suburban route
Urban-suburban route
Urban-suburban route
Urban route
Urban route
Light-rail line, urban route
Light-rail line, urban route
a Ranges are for averages For runs along each route. bMean stops per mile rounded to nearest integer. cExcludes terminal stops. dExcludes main CBD stops.
4 Transportation Research Record 915
Table 6. Typical dwell times at major bus stops, 1979-1980.
Observed Dwell Time (sec)
Type of Stop City Route Location Time of Day Mean SD
Ent! uf bus line al rail Boston I , Massachuseus Avenue Harvard Square, Ked Lme p.m., midday 33 18 transit station Dudley Square, Orange Line p.m., midd ay 38 13
71, Watertown Brattle Station, Red Line a.m. 37 NA 240, Randolph Ashmont, Red Line a.m. 55 NA
Transfer point at rail Boston I , Massachuse tts Avenue Auditorium, Green Line p.m. 36 8 transit station 1, Massachusetts Avenue Central Square, Red Line p.m. 24 6
77, Arlington Ltd. Harvard Square, Red Line a.m., midday 25 NA Chicago 11, Lincoln Western, Ravenswood a.m. 23 NA
11, Lincoln Fullerton, North, South a.rn. 23 NA
Major transfer point Chicago II, Lincoln Foster a.m. 40 NA to another bus line New Haven Congress, Savin Rock West Haven Center p.m. 29 9
Major non-CBD stop, Boston 71, Watertown Watertown Square a.m., midday 26 4 movie, town hall, Boston 77, Arlington Ltd. Three major stops p.m. 34 6 hospital, school, etc. New Haven Congress Yale, New Haven Hospital p.m. 40 18
Notes: NA= not available. Data are from field studies.
Table 7. Designated versus actual bus stops, 1979-1980.
Time of No. of City Route Day Runs Direction
Boston 1, Massachusetts Avenue p.m. 2 NB 71, Watert1>wn-Brattle a.m. 2 EB 240, Randolph-Ashmont a.m. 2 NB 1, Massachusetts Avenue Midday 4 NB, SB 71, Watertown-Brattle Midday 4 EB, WB 77, Arlington Heights Ltd. Midday 4 NB, SB 240, Randolph-Ashmont Midday 2 NB, SB
Chicago 11 , Lincoln a.m. l SB 11, Lincoln, heavy I. 2 miles a.m. I SB 22, Clark Midday I SB
New Haven Q, Edgewood a.m. 3 EB
Note: Data are from field studies.
know how often a bus stops as well as how long. Table 7 compares the number of sch,eduled stopping places with the bus stops actually made during peak and off- peak conditions. During peak hours local buses stopped at 68 to 78 percent of the designated stopping places. During off-.pP.llk perioils, the ratio of actual to scheduled stops was as low as 30 percent. These figures suggest that transit systems could reduce the number of designated stops without adversely affecting ridership.
ACCELERATION AND DECELERATION TIME
Bus acceleration and deceleration time was computed by two separate methods, and the results were then compared. Actual times observed in field studies were summarized. Times were computed based on assumed cruise speeds and rates of acceleration and deceleration set forth in the first edition of the Transportation and Traffic gineering Handbook (!!J . In effect, speed pr ofiles were deve loped for various stop spacings.
Table 8 Cl,.!.£) gives detailed data on bus acceleration and deceleration based on various field studies. Total acceleration and deceleration time per stop ranged from 11 to 23 sec, depending on stop frequencies. Analysis of these data showed that the total acceleration plus deceleration time per stop followed this formula:
T = 23. 4 - l. 53X Rs -0.78
Stops per Mile
Actual to Scheduled
Scheduled Actual (%)
6 .5 7.6 1.7 6.5 7.6 4.7 1.7
7 .7 8 .3
11.2
6.6
5.2 80.0 6.7 88.0 1.5 83.3 3.4 52.3 3. 1 40.8 1.4 29.8 1.3 76.7
5.2 67 .5 6.7 80.7 6.5 58.0
5.7 85 .8
where X is the number of stops made per mile and T is the total acceleration and deceleration time per stop.
Acceleration and deceleration time based on this formula is compared below with that obtained based on theorP.t.iC'!nl ca l culations: The theoretical calculations assumed that a bus accelerates at its normal or maximum rate to reach the maximum possible cruise speed and subsequently decelerates at the maximum comfortable rate to a full stop (]:) :
Acceleration and Oece l e r at i on Time ! s ec !
No. of Stops Field Theoretical Eer Mile Survey Calculation l 21.9 44-62 2 20.3 44 3 18.8 37 4 17.3 30 5 15.8 24 6 14. 2 24 7 12.7 18 8 11.2 13-18
The acceleration and deceleration time observed in the field was consistently less than that derived from the vehicle performance calculations, especially as the spacing between stops increases. For example, at six stops per mile, the field surveys found a 14-sec acceleration and deceleration time, whereas a bus
Transportation Research Record 915 5
Table 8. Observed bus acceleration and deceleration times per stop.
Acceleration and Deceleration Time per Stop
Avg Stops (T) (sec) per Mile
City Route (X) Mean SD Remarks
Boston J 3 14.8 3.8 Urban, high-density line 71 3 20.2 4.3 Suburban line 77 I 22.3 5.7 Suburban, limited stops 240A I 22.5 5.0 Suburban
Oakland 51 NA 17.3 3.0 Suburban New Haven Q 6 10.8 1.8 Urban Chicago II 5 18.0 2.7 Urban
22 6 16.8 5.9 Urban, high density Hong Kong Small single-deck bus 14.5 4.4
Large single-deck bus 16.7 4.4
Note: Data are from field studies.
Ta.hie 9. Bus travel times and speeds as function of stop spacing.
Dwell Time per Stop
IO sec 15 sec 20 sec 30 sec
Stops Minutes Miles Minutes Miles Minutes Miles Minutes Miles per Mile per Mile per Hour per Mile per Hour per Mile per Hour per Mile per Hour
I 1.97 30.5 2.05 29.3 2.13 28.2 2.30 26.1 2 2.40 25 .0 2.56 23.4 2.73 22.0 3.07 19.5 3 2.85 21.0 3.10 19.4 3.35 17.9 3.85 15.6 4 3.27 18.3 3.60 16.6 3.93 15.3 4.60 13.0 5 3.75 16.0 4.17 14.4 4.58 13.l 5.42 I I. I 6 4.30 14.0 4.80 12.5 5.30 11.3 6.30 9.5 7 4.67 12.8 5.25 11.4 5.83 10.3 7.00 8.6 8 5.33 11.3 6.00 10.0 6.67 9.0 8.00 7.5 9 6.00 10.0 6.75 8.9 7.50 8.0 9.00 6.7
IO 7 .00 8.6 7.83 7.7 8.67 6.9 10.33 5.8 12 8.17 7.3 9.23 6.5 9.33 6.4 11.33 5.3
Note: Based on an acceleraHon and declerntion rate o f 3 mph/sec and acceleration-deceleration times o bse rved in field.
Figure 3. Bus speed versus stops.
50 :i:: ~ 40 I
fil 30 w 3> 20
10
0
~ ""' 10•1 STO~
....... ....;:: ....... ----30• ST< IP"'"
0 2341587 STOPS PER MILE
8 SI 10
reaching its maximum possible cruise speed would spend 24 sec accelerating and decelerating.
Several factors underlie these differences:
1. Buses usually do not reach their maximum attainable cruise speeds between stops when operating on cit:S,- streets because of posted speed limits, intersection interference, traffic signal controls, or street congestion. A bus making one stop per mile on a suburban street may never exceed 30 to 35 mph even though theoretically it would reach 50 to 60 mph.
2. Acceleration sometimes takes place through a series of steps in which the bus operates at several cruise speeds. Only the first step was considered as acceleration in the field.
Bus travel times and speeds as a function of stop
spacing, dwell times, and observed acceleration and deceleration patterns are shown in Figure 3 and Table 9. Bus speeds as a function of stop spacing are similar to those reported in previous studies (13). These exhibits provide a practical guide for estimating bus travel times for various operating conditions and for assessing the changes in travel times resulting from reducing the frequency and duration of stops .
For example, at eight stops per mile and 20 sec/ stop, bus travel time is 6 min. If the stops are reduced to six per mile and the dwell time to 10 sec/ stop, bus travel time would be 4.30 min. This time saving exceeds the minute per mile buses normally lose due to traffic delay.
APPLICATIONS
General guidelines for peak-hour bus dwell times and stop frequencies as a function of location and route type are summarized below. These data provided inputs for Table 9 in estimating overall bus performance.
Passenger stops made per mile, passenger dwell time per stop, and acceleration and deceleration time per st.op are given as a function of general location:
1. The number of passenger stops per mile actually made decreases with decreasing population density; suggested values are B, CBD; 6, city; 4, inner suburbs; and 2, outer rural areas.
2. Passenger dwell times (seconds per stop) range from 30 (average) to 60 (major) sec in the CBD; they
6
average 15 sec in the city and 10 sec/stop in suburban areas.
3. Acceleration and deceleration time loss per stop average 11-13 sec in the CBD, 14-15 sec in the city, and 17-25 sec in suburban areas.
The type of route and type of stop vary among urban areas; they reflect ridership densities (reported by the transit agepcy), route configuration, a nd land use patterns.
Type of Route
Suggested guidelines for bus dwell times by type of route are given next (these exclude the CBD). A heavy urban route, for example, would have stops averaging 20 sec as compared with 16 sec for a medium route and 12 sec for a light route. For suburban and rural areas, heavy routes would have stops averaging 16 sec; medium routes, 12 sec; and light routes, 8 sec .
CBD Stops
Guidelines for peak-hour dwell times at CBD bus stops are shown below (based on 1979 New Haven data):
Peak-Hour Dwell Time \ s ee) T:r:Ee of StoE Maximum Avg . Mean SD Business 120 50 ~ Other 60 . 20 15 Outlying 20 10 7
Bus dwell times will average 50-60 sec at the busiest stops, 20-30 sec at most stops, and 10 sec at lightly used stops on the CBD fring e . The maximum dwell times will be twice these values.
Major Bus Sto12s
Suggested guidelines for dwell times at major bus stops during evening peak hours include 40 sec for the end of the bus line at rail transit, 30-35 sec at the transfer point to rail transit or at a major bus stop, and 30-35 sec at other major stops.
Sto12s 12er Mile
Guidelines for the number of bus stops per mile actually made by type of route and area are given below:
Type of Bus Sto12s 12er Mile Route Urban Suburban Rural Heavy 7 5 3 Medium 6 4 2 Light 5 3 2
Buses operating on a heavy urban route would make seven stops per mile as compared with six for a medium urban route and five for a light one.
IMPLICATIONS
The preceding parameters and relationships can be used directly in developing and assessing operating and service changes. They also provide inputs to long-range planning procedures. Field studies should be conducted to obtain city-specific parameters if greater precision is needed .
Several service planning and policy implications are apparent. Transit performance should be improved by keeping the number of stopping places to a minimum. Fare-collection policies and door conf igura t ions and wi dths are espec ially important in r educing dwell times along high-density routes. Many
Transportation Research Record 915
European transit systems have adopted such actions, but implementation in the United States generally has been limited even though the U.S. transit industry has recognized the need for fewer stopping places for 75 years.
It is desirable to eliminate traffic-induced congest ion by improving general traffic flow or by providing bus priority lanes or streets or, in selected situations, bus signal preemption. These actions will improve bus performance in congested areas. Nevertheless, these gains often may be less than those resulting from reducing passenger service delays over the entire system. Herein lies an important challenge to transit operators.
ACKNOWLEDGMENT
The research summarized here was conducted for UMTA in 1980. The help of Thomas Hillegass and Larry Quillian of UMTA and Herbert Burstein of the Regional Planning Agency of South Central Connecticut is especially appreciated.
REFERENCES
1. H.S. Levinson. !NET Transit Travel Time Analysis. UMTA, Final Rept., April 1982.
2. R.B. Dial, G.S. Rutherford, and L. Quillian. Transit Network Analysis: !NET. UMTA, July 1979.
3. Study and Recommendations for Improving Traffic Movement in the Central Business District. City of Chicago, 1950.
4. Dallas Bus Operational Study, Volume 2. Wilbur smith and Associates, New Haven, CT, 1972.
5. Transit Development Program, New Haven, Connecticut. Wilbur Smith and Associates, New Haven, CT, 1975.
6. San Jose--Santa Clara County Bus Study. Wilbur Smith and Associates, New Haven , CT, 1968.
7. Alameda--Contra Costa Transit District Bus Priority Techniques Study. California Department of Transportation, Sacramento, draft final rept., June 29, 1979 .
8. R. Edminster and D. Koffman. Streets for Pedestrians and Transit: An Evaluation of Three Transit Malls in the United States. UMTA, Rept. UMTA-MA-06-0049-79-1, Feb. 1979.
9. St. Louis Metropolitan Area Transportation Study. w.c. Gilman and Co., St. Louis, MO, 1959.
10. Surface Car Line Operations Study: Street, Green Line, Brookline, and Bruce Campbell and Associates, Boston, 1969.
Beacon Boston. MA, May
11. Transportation and Traffic Engineering Handbook, 1st ed. (J. Baerwald, ed.). Prentice-Hall, Englewood Cliffs, NJ, 1976.
12. H.S.W. Leong. Warrant for Provisions of a · Bus Bay. Proc., Australian Road Research ,Board, Vol. 4, No. 1, 1968.
13. Bus Route and Schedule Guidelines. NCHRP, Synthesis 69, 1980.
Publication of this paper sponsored by Committee on Bus Transit Systems.